Temporal and spatial variability of distributed PV impacts on Australian distribution network load profiles

Transcription

1 Temporal and spatial variability of distributed PV impacts on Australian distribution network load profiles Navid Haghdadi 1,2,3, Rob Passey 1,2,3, Anna Bruce 1,2, and Iain MacGill 2,3 1 School of Photovoltaic and Renewable Energy Engineering, UNSW 2 Centre for Energy and Environmental Markets (CEEM), UNSW ( ) 3 School of Electrical Engineering and Telecommunications, UNSW CIGRE Australia. Slide 1

2 Context More than 80% of the 6GW of PV installed in Australia is comprised of distributed PV systems installed on residential, commercial, and industrial rooftops. Number of installations % of Dwellings 0 QLD SA WA VIC NSW ACT TAS NT 0 installations percentage_dwellings Australian PV Institute (APVI) Solar Map (as of April 2016) CIGRE Australia. Slide 2

3 Context Unlike conventional utility-scale generation, small-scale PV systems are rarely systematically monitored Thus, difficult to investigate the performance and hence impact of them on the network The impact also depends on the underlying network load profile, which varies across different climates and according to the spatial characteristics of both PV and energy consumers across the network Reneweconomy CIGRE Australia. Slide 3

4 PV performance in DNSPs Therefore, we have used historical generation data from thousands of distributed PV systems in Australia to estimate the average performance of distributed PV in DNSPs zone substations (ZS). The DNSPs include (total 518 ZS): Ausgrid in NSW (~170) ActewAGL in ACT (10) Energex in QLD (237) Powercor (2), Ausnet (52), Citpower (2), and United Energy (45) in VIC AEMO CIGRE Australia. Slide 4

5 Method Average PV performance is estimated in each 2-digit postcode area for each half hour from [1] PV performance is found for the location of each ZS Peak load PV performance for each ZS is extracted for analysis Note that the result would have been more accurate if we could have the actual capacity of the installed PV and hence underlying load in each ZS. [1] The method is used in APVI live map and described in: Haghdadi et al. Assessing the representativeness of Live distributed PV data for upscaled PV generation estimates. Power and Energy Engineering Conference (APPEEC), IEEE PES Asia-Pacific, November 2015, Brisbane, Australia CIGRE Australia. Slide 5

6 Method CIGRE Australia. Slide 6

7 Method CIGRE Australia. Slide 7

8 Method CIGRE Australia. Slide 8

9 Method CIGRE Australia. Slide 9

10 Method CIGRE Australia. 10 Slide

11 Method CIGRE Australia. Slide 11

12 Results Average PV performance in top 1% of peak times in each zone substation CIGRE Australia. Slide 12

13 Result Average PV performance in top 1% of peak times in each year Ausgrid Ausnet CIGRE Australia. Slide 13

14 Result Impact of load type Energex United Energy CIGRE Australia. Slide 14

15 Result Assured level of PV generation in peak times CIGRE Australia. Slide 15

16 Conclusions The correlation of PV performance and peak load has been assessed in different zone substations in Australia The performance of distributed PV and hence its potential for peak reduction is found to be variable in different locations (most probably due to variable climate and hence load profile) The impact can vary between different years (most probably due to the inter-annual variability of load and solar potential) The minimum assured level of PV varies between different load types CIGRE Australia. Slide 16

17 Thank you! CIGRE Australia. Slide 17